Essays In The Economics Of Violence And Conflict

Chapter 1 studies the impact of the arrival of displaced individuals on the wages of residents at receiving locations. To do this, the study employs an enclave IV strategy, which exploits both social distance between origin and destination locations and forced migration. I compare the effects on four different subgroups of the population, partitioned by skill (low-skilled versus high-skilled) and by gender. The analysis suggests that a conflict-induced increase in population leads to a short-run negative impact on wages, but that subsequent out-migration from receiving municipalities helps to mitigate these effects. Chapter 2 studies the relationship between political violence and congressional decision-making. I examine how politicians and their constituents respond to attacks by FARC, Colombia's largest rebel group, using data from politicians' Twitter accounts and roll-call voting records, and employing both an event study and a difference-in-differences research design. The analysis finds that tweets from incumbent politicians and tweets which exhibit "right-wing" language receive higher user engagement (a proxy for popular support) following rebel attacks. In congress, politicians were more likely to align their legislative votes with the right-leaning ruling party following an attack, before the government started negotiations with the rebels in 2012. However, this relationship breaks down after the start of the peace process. The empirical results are consistent with a political economy model of legislative behaviour in which events that shift median voter preferences, and the presence of rally 'round the flag effects, elicit different politician responses depending on the policy position of the ruling party. Chapter 3, joint work with Gustavo Bobonis and Roberto Castro, provides evidence of the long-term relationship between male-to-female spousal violence and the Oportunidades conditional cash transfer program in rural Mexico. It uses data from three nationally representative surveys that include detailed information on the prevalence of spousal abuse and threats of violence against women. Constructing comparable groups of beneficiary and nonbeneficiary households within each village to minimize potential selection biases, the analysis finds that, in contrast to short-run estimates, physical and emotional abuse rates over the long term do not differ significantly between existing beneficiary and nonbeneficiary couples.

Chapters 1 and 2 of my thesis deal with long-term economic effects of violent conflicts. In the first chapter, I use World War II casualties suffered in Austrian municipalities as a natural experiment for human casualties and find a significant negative causal effect of human losses on economic activity today. As I demonstrate, the likely channel through which the effect persisted over time is through its impact on the structural composition of the work force. Specifically, greater human losses increased the fraction of employment in manufacturing at the expense of agriculture until the 1970s and services from then onwards. A simple model shows that structural change can translate a lower labor share in agricultural production into less participation of service sector growth at a later time. In the next chapter, I identify a channel through which the disadvantage of displacement during a violent conflict might be carried over to the next generation. In particular, I show that displaced parents spend significantly less on the education of their children years later. A decomposition of the causal effect shows that differences in income and the stock of durable goods can at most explain one third of the finding. Some evidence points towards increased uncertainty about the future of displaced parents and hence reduced spending on non-vital expenditure positions. The final chaper revisits the paper by Algan & Cahuc (AER, 2010) in which they find that inherited trust has a large impact on GDP per capita. First, I show that the estimates presented in Algan & Cahuc might be biased due to a difference between the lag structure of inherited trust and initial income in their econometric specification. Next, I focus on their robustness checks, where I replicate their results and document that most of their robustness checks fail when a programming error and data problems are corrected. I conclude that their results should be considered with great care.

This dissertation consists of three chapters on topics relating to conflict, social cooperation and development economics. Several studies have identified the impact of adverse economic shocks on civil conflict using rainfall variation as an instrument for income or growth. The first chapter contributes to this literature by carrying out a micro-level analysis on the relationship between climate and resource variation with armed conflict using a novel dataset on inter-tribal violence manifested through livestock raids in a pastoral-dependent region of East African called the Karamoja. Consistent with previous work, I find that across the region there is a negative relationship between resources and conflict, when resources are measured with forage. However, I also find that both decreases and increases in rainfall are correlated with conflict across the region. This bimodal relationship between precipitation and conflict persists when I analyse raid-location and tribe specific variation in rainfall, while the relationship between forage and raiding is less clear. There is some indication that forage-scarcity motivates tribes to carry out raids and forage-scarce sublocations appear to be more vulnerable to raids and livestock losses, but these results are not robust to all specifications. In the second chapter, I study the effect of Uganda's 2006 disarmament policy in the Karamoja region in East Africa. The disarmament policy greatly reduced the guns of tribes in the Ugandan districts of the region but not in the Kenyan districts. The theoretical impact of the disarmament is ambiguous, however, since guns can be used for deterrence as well as helping aggressors carry out violent crimes, such as livestock raiding. Empirically, I find that the disarmament campaign had the unintended effect of increasing the frequency of raids in Uganda by about 40%, while, consistent with the idea that disarmament reduced the costs of raiding, I find no impact on the monthly death rate. Moreover, the increase in raids in Uganda was driven by an increase in Ugandan initiated raids on other Ugandans, not an increase in Kenyan initiated raids on Ugandans, suggesting that within Uganda the deterrent effect of guns outweighs their impact as a tool of aggression. In the third chapter, written jointly with Johannes Haushofer, we study the impact of stress on social behavior by exogenously stimulating the two biological systems associated with stress: the hypothalamus-pituitary-adrenal axis (HPA) and noradrenergic (NA) system and measuring behavior in interactive tasks in a laboratory experiment. Our preliminary findings suggest that the concurrent stimulation of both systems, through the administration of 60mg of hydrocortisone and 20mg of yohimbine, did not lead to statistically detectable changes to behavior in any of the social tasks. It did, however, manifest in lower opinions of the trustworthiness and fairness of other people, as well as a decrease in the value associated with helping other people, as measured through a visual analog scale survey. Given these initial results, we find preliminary evidence for a relationship between stress and anti-social behavior as revealed through lower beliefs on social standards. JEL Classification: C91, K42, Q56